The automation of lifetime value: How developers can track LTV

A consumer finds an app, downloads it, uses it once and then never again. It's not hard to calculate the lifetime value of that sort of user. For those that actually engage with apps and mobile games, however, it can get a lot more complicated. 

Amobee LTV is designed to offer a holistic view of what users are doing with multiple apps.

Lifetime value (LTV) can be defined in many ways, but for the most part it's a determination of how much revenue each customer will generate based on the length of time they use a particular app. As discussed previously on FierceDeveloper, working with LTV takes some getting used to, but it can be extremely helpful in letting developers be more strategic with the way they market their apps and how they attempt to monetize them. That may be why more vendors are creating dashboard-type interfaces to automate the process of tracking and optimizing mobile app LTV. 

A few weeks ago, for example, Redwood City, Calif.-based Amobee announced Amobee LTV, which is designed to offer a holistic view of what users are doing with multiple apps, including the ability to segment audiences and generate a lifetime value score for each app or customer segment. For example, developers could configure the platform so that it assigns a low LTV score for someone who only uses an app or plays a mobile game once a month, or higher if they use it several times a day and make in-app purchases. Those with high LTV scores might be good candidates for promotions about other merchandise or offers, while those with low LTV might benefit from a rich-media push notification, which can also be generated through the Amobee product.

Keep it simple 
According to Gil Sheinfeld, Amobee's CTO, the system will fill a gaping hole in the amount of information many developers need to know if their apps are not only finding an audience but keeping them. 


"They're practically blind. They don't know about the actual results. These are all key," he said.

Although there are a number of broad mobile analytics tools already available for developers that may incorporate LTV data, Sheinfeld said developers are demanding less onerous ways of getting at it.

"They're complex, they're not intuitive," he said. "We think that that is one of the key advantages of the platform--is that information is there in a simple way. No need for complex information flows."

Amobee is far from the only competitor in this segment, however. Sandpoint, Idaho-based Kochava also recently released version 2.0 of its mobile app attribution tracking platform. The company said its system integrates with more than 300 mobile ad networks, publishers and exchanges, including Facebook, Google, Twitter, Yahoo and Pandora. It is also designed to track users and campaigns based on device, platform and app versions to number of sessions, country and more.

According to Kochava CEO Charles Manning, customer understanding of LTV is maturing quickly, but some education is still necessary--not just with developers, but even brands and agencies. 

"These prospects often ask if we can 'track LTV.'  The quick answer is yes," he said. "The more involved conversation surrounds the definition of 'success' and 'value.'" 

In free to play games, for instance, Manning said many customers analyze D1 (Day 1), D3, D7 and D14 metrics with cohorts of users. What this means is that the advertiser is identifying key metrics such as registration, tutorial completion, monetization through in-app purchases and so on. These metrics are then incorporated into a funnel. As cohorts of users pass through these timelines, advertisers that see certain things that happen by D1, like registration and tutorial completion, then translate the probability of the revenue that they will produce by D3. This could be one way of showing success with a particular campaign based on LTV tracking. 

Kochava says its system integrates with more than 300 mobile ad networks, publishers and exchanges.

"If [the] metrics are proving to be meaningful, then advertisers will increase their spend with those media sources that brought that cohort," Mannning said. "The goal may not be in-app purchases. They could be brand engagement or viral shares."

Fraud skews scoring
Even with the right metrics and the most intuitive dashboards, though, LTV can sometimes be tricky. Ryan Greiner, vice president of product development at San Francisco-based Apsalar, notes that if someone uses a jailbroken phone to create a phony receipt of an in-app purchase, they could make some LTV scoring highly inaccurate. 

"If you'e a developer and have a whole bunch of things you're tracking, one of the last things you're probably looking at it is in-app transaction fraud," he said, noting some developers have seen fraud rates as high as 50 percent. Greiner said there have been changes in iOS 7 and other platforms that are helping reduce in-app purchase fraud rates, but it shows that true LTV can be elusive. 

Manning said developers also need to think about how they feel about where their LTV data could end up. 

"The market is broken out into two categories of vendors: those that collect data and use it for other business-driven purposes--usually to sell or leverage when selling media--and those who are truly independent and only collect data for the purpose of serving the company who owns the data," he said, adding that the former model explains why some services are free or offered at discounted rates. "Kochava has been successful in supporting those customers who want to keep their data their own."

As they choose the right platform for tracking lifetime value of a mobile app user, developers will have to weigh these different business models and invest with whoever will serve them best. After all, to the vendors in this space, developers probably all have their own LTV scores, too.